Douglas Kronaizl

Douglas Kronaizl is a staff writer at Ballotpedia. Contact us at

Four Pivot Counties flip from Trump to Biden as results continue to be updated

Ballotpedia has been analyzing the 206 Pivot Counties that voted for Barack Obama (D) in 2008 and 2012 and Donald Trump (R) in 2016 by creating two new categories: Retained Pivot Counties, which voted for Trump again in 2020, and Boomerang Pivot Counties, which voted for Joe Biden (D). The analysis continues to shift as states certify their election results.

Since publishing our initial Pivot County analysis, four Retained Pivot Counties have flipped to become Boomerang Pivot Counties. Additionally, ten new counties have released vote totals, resulting in nine new Retained Pivot Counties and one new Boomerang Pivot County.

There are currently 179 Retained Pivot Counties and 25 Boomerang Pivot Counties. These numbers are still subject to change.

The four counties that flipped from Retained to Boomerang are all located in New York: Broome, Essex, Rensselaer, and Saratoga.

Based on current results, Biden’s margins of victory in Broome, Essex, and Rensselaer are lower than Obama’s in 2012, the last time a Democrat won these counties. Biden exceeded Obama’s 2012 margin in Saratoga County. The table below shows the unofficial results with vote totals in parentheses.

Biden also won Kennebec County, Maine, by a margin of 0.39 percentage points, less than Obama’s 2012 margin of 12.85 percentage points,

Trump won nine new Retained Pivot Counties, two in Mississippi and seven in Maine. Compared to 2016 results, his margins of victory increased in four and decreased in five. Those counties are listed below, split into those where his margin increased and those where it decreased:

Two counties—Alexander and Henderson, Illinois—have not yet released results.

Ballotpedia will continue to provide updates as results become available. For more information updated weekly, click here.

A closer look at historical margins of victory in Boomerang and Retained Pivot Counties

Following the 2016 presidential election, there were 206 Pivot Counties that voted for Barack Obama (D) in 2008 and 2012 before voting for Donald Trump (R) in 2016.

Based on unofficial 2020 results, there were 22 Boomerang Pivot Counties, which flipped to Joe Biden (D), and 174 Retained Pivot Counties, which voted for Trump again.

In 2016, Trump’s average margin of victory in what are now the 22 Boomerang Pivot Counties counties was 1.97 percentage points, 9.45 points fewer than his average across all Pivot Counties. By contrast, Obama overperformed in these counties, relative to his overall averages, in 2012 and 2008.

On the other hand, compared to 2016, Trump’s average margin of victory in the 174 Retained Pivot Counties increased by 2.15 percentage points. Trump had overperformed in the Retained Pivot Counties in 2016, exceeding his overall average margin of victory by 1.06 percentage points. Obama underperformed in these counties in 2008 and 2012.

To learn more about Boomerang Pivot Counties, click here:,_2020:_Boomerang_Pivot_Counties

To learn more about Retained Pivot Counties, click here:,_2020:_Retained_Pivot_Counties

Former Ohio House Speaker Larry Householder wins re-election

Incumbent Larry Householder (R) defeated four write-in candidates—Marci McCaulay (D), Jay Conrad (R), Robert Leist (L), and Kaitlyn Clark (I)—in the general election for Ohio’s House of Representatives District 72. Householder ran unopposed in the April 28 Republican primary.

On July 21, 2020, after the filing deadline for additional candidates to appear on the ballot, Householder was arrested and charged with conspiracy to participate in a racketeering scheme. He was accused of collecting more than $60 million in exchange for legislation that would bail out two nuclear plants in Ohio. The bailout was valued at $1.5 billion.

At the time of his arrest, Householder had been serving as Speaker of the House since 2019 when he defeated sitting House Speaker Ryan Smith (R). The House removed him from the position by a 90-0 vote on July 30.

In response to his arrest, on Sept. 1, Householder said, “[I]n the United States, we believe that you’re innocent until you’re proven guilty. And that day has not occurred … And so, I am innocent. I am going to defend myself vigorously.” Householder entered a plea of not guilty before U.S. Magistrate Judge Stephanie Bowman on Sept. 3.

Householder previously served in the House from 1997 to 2004, including a tenure as Speaker of the House from 2001 to 2004. He returned to the House representing District 72 in 2016 and won re-election in 2018.

Analyzing margins of victory in the 206 Pivot Counties nationwide

Voters in 206 Pivot Counties across the country backed Barack Obama (D) in 2008 and 2012 and Donald Trump (R) in 2016. How did these counties vote in 2020?

We have split the Pivot Counties into two categories based on the unofficial results: 

  1. Retained Pivot Counties, which voted for Trump again in 2020, and 
  2. Carousel Pivot Counties, which voted for Joe Biden (D) this cycle.

Preliminary analysis shows the following breakdown for the 206 Pivot Counties:

  1. 174 Retained Pivot Counties
  2. 20 Carousel Pivot Counties
  3. 12 unclear/too-close-to-call

Trump has been winning the 174 Retained Pivot Counties with an average margin of victory of 14.9 percentage points. Compared to his 2016 results, Trump’s margin of victory decreased in 49 Retained Pivot Counties and increased in 125.

Biden has been winning the 20 Carousel Pivot Counties with an average margin of victory of 3.1 percentage points. Compared to Obama’s results in 2012, the last time a Democrat won in these counties, Biden’s margin of victory represents a decrease in 18 and an increase in two.

The five counties with the largest change in margin of victory for Democrats and Republicans since 2016 are in the tables below.



Revisiting the two presidential election recounts in 2016

Following the 2016 presidential election, two states—Wisconsin and Nevada—conducted recounts after receiving requests from candidates. Neither recount changed the election outcome. Below is a brief look at those recounts, who requested them, and what effect they had on vote totals


Green Party candidate Jill Stein requested a full recount in Wisconsin on Nov. 25, saying the election had been hacked. Prior to the recount, Donald Trump (R) led Hillary Clinton (D) by 27,257 votes. The recount began on Dec. 1 and finished on Dec. 12. As a result, Clinton gained 713 votes and Trump gained 844, adding 131 votes to his margin of victory.


Partly in response to Stein’s requested recount in Wisconsin, Reform Party candidate Rocky De La Fuente requested a partial recount of four counties and Carson City, Nevada, on Nov. 29. The recount began on Dec. 5 and finished on Dec. 8. As a result, Trump lost six votes and Clinton lost nine, subtracting three votes from her margin of victory.

Stein also requested recounts in Michigan and Pennsylvania but neither was completed. In Michigan, Stein ended her request after a series of court challenges, which involved a state ruling that Stein had no standing to request a recount. In Pennsylvania, Stein withdrew her request amid additional court challenges.

According to a study by FairVote, between 2000 and 2015, there were 4,687 statewide general elections, 27 of which (0.58%) resulted in statewide recounts. Of those 27, three changed the election outcome: Minnesota’s 2008 U.S. Senate election, Vermont’s 2006 State Auditor election, and Washington’s 2004 gubernatorial election. None of those three swung the winning candidate by more than 500 votes. FairVote describes itself as a “nonpartisan champion of electoral reforms that give voters greater choice, a stronger voice, and a representative democracy that works for all Americans.”

22 of Ballotpedia’s 57 federal battleground elections taking place in states with automatic recount procedures

Image of donkey and elephant to symbolize the Democratic and Republican parties.

This November, 22 of the 57 U.S. House and Senate races Ballotpedia identified as battlegrounds are taking place in states where a close vote could automatically trigger a recount under state law.

An automatic recount occurs if election results meet certain criteria laid out in state law. The most common trigger for an automatic recount is when election results fall within a close vote margin.

The table below lists those battleground races alongside the most recent margin as well as the state’s close vote trigger.

For seven of the races listed below, an automatic recount is only triggered if the race ends in a tie. For the remaining fourteen races, a recount is triggered if the vote total falls within the given close vote margin.

Nationwide, there are twenty-three states where a close or tie vote could trigger an automatic recount. However, a close vote is not the only way a state might trigger an automatic recount. Four states—Kentucky, Minnesota, New York, and North Carolina—require an automatic recount if an error is discovered while tabulating the votes or while conducting a post-election procedure such as an audit of voting machines. The map below shows those states with policies requiring an automatic recount under certain circumstances.

Additional reading:

What happened in two redo elections in 2020?

In 2020, Ballotpedia has identified two elections where the results have been voided and a redo election scheduled in their place: a city council election in Paterson, New Jersey, and a sheriff election in Iron County, Missouri.

A redo election, also known as a revote or special election remedy, is the process of voiding election results and holding a new election. The specific reasons for calling a redo election vary, but might include deliberate efforts to obscure the results such as electoral fraud or mistakes like a broken voting machine.

Typically, states or courts call a redo election only after an interested party—normally a candidate, voter, or election official—contests the election results. While most states have provisions describing how to handle contested elections, these provisions do not normally specify what to do if fraud or mistakes occur.

Most redo elections, like the two described below, take place at the municipal or county level. The most recent redo election for a federal office took place in 2019 in North Carolina’s 9th Congressional District. The last federal redo election before that was in 1974.

Paterson, N.J., city council election

Five candidates—incumbent councilman William McKoy, Chauncey Brown, Sharrieff Bugg, Alex Mendez, and Robyn Spencer—ran in the May 12 city council election for the 3rd Ward in Paterson, New Jersey. Initial results showed Mendez defeating McKoy with 1,595 votes to McKoy’s 1,350, a 245 vote margin. A recount narrowed the margin to 240 votes. Election officials conducted the election entirely by-mail due to the coronavirus pandemic.

On June 14, McKoy contested the election results alleging absentee/mail-in electoral fraud in the form of ballots submitted on behalf of voters who later alleged they never received absentee/mail-in ballots. During the May 12 election, election officials rejected 24% of absentee/mail-in ballots in the 3rd Ward compared to 10% s.

On Aug. 16, Passaic County Superior Court Judge Ernest Caposela voided the May 12 election and ordered a redo election for Nov. 3, 2020.

Iron County, Mo., sheriff election

Incumbent Roger Medley, Ryan Burkett, Brian Matthiesen, Ben Starnes, and James Womble participated in the Aug. 4 Republican primary for sheriff in Iron County, Missouri. Burkett defeated Medley by 73 votes.

Medley contested the election, alleging the usage of incorrect ballots, a voting machine missing part of its tally tape, and violations of state law such as the mother-in-law of one candidate working as an election judge.

On Aug. 27, Iron County Circuit Judge Kelly Parker voided the election results and set a redo primary election for Sept. 22. Burkett defeated Medley in the redo primary election receiving 42% of the vote to Medley’s 27%.

Share of incumbents defeated in contested primaries grows for third even-year cycle in a row

Ballotpedia’s annual state legislative competitiveness study shows that for the third even-year election cycle in a row, the share of incumbents defeated in contested primaries has grown. In the 44 states that held state legislative elections this year, 153 incumbents—61 Democrats and 93 Republicans—were defeated by primary challengers.

Overall, 15.2% of the 1,016 major-party incumbents who faced primary challengers in 2020 lost, the third consecutive increase compared to 2018 (13.8%) and 2016 (12.3%). The loss rate in 2020 also exceeded that of 13.0% in 2014.


  1. 18% of Republicans who faced challenges in 2020 were defeated—the highest since at least 2014.
  2. 12% of Democrats who faced challenges were defeated, lower than the 14% rate in 2018.
  3. More Democrats were defeated in states with Democratic trifectas, as was the case for Republicans in Republican trifectas.
  4. The loss rate for incumbents in states with divided governments exceeds the national average altogether and by party.
  5. Democratic incumbents were defeated at the highest rate in states with Democratic trifectas and at the lowest in those with Republican trifectas. In states with divided governments, the rate exceeded the national average for Democrats.
  6. Republican incumbents facing contested primaries were more likely to be defeated in states with divided governments than in states with trifectas.

To read more about the state legislative incumbents defeated in primaries this year, click here:

Percent of U.S. House races contested by both major parties reaches a century high

More than 95% of elections for U.S. House (415 of 435) in 2020 are contested by candidates from both the Democratic and Republican parties, according to Ballotpedia’s Annual Congressional Competitiveness Report. This is an increase over the four preceding election cycles and the highest percentage of contested races for U.S. House since at least 1920.

Twenty races (4.6%) feature no major party competition. Democrats are running in twelve races without Republican competition and Republicans are running in eight without Democratic competition.

The presence of major party competition has increased in each election cycle since 2014 before reaching this century high. In 2014, 16.8% of races (73) had no major party competition, meaning 83.2% (362) were contested by both major parties. The percentage of contested U.S. House races increased to 87.8% (382) in 2016 and 91.3% (397) in 2018.

A Democratic candidate is running in 98.2% of House races (427 of 435), down from 99.5% (433 of 435) in 2018.

A Republican candidate is running in 97.2% of House races (423 of 435), up from 91.7% (399 of 435) in 2018, and the highest rate of Republican competition since at least 2012.

All 33 regular U.S. Senate races in 2020 have major party competition. The most recent U.S. senator to run without major party competition was Jeff Sessions (R-Ala.), who ran unopposed in the 2014 general election.

To read more about major party competition, open seats, and primary competitiveness in congressional elections, click here:,_2020

Kansas’ state legislative Republican primaries alter the ideological makeup of GOP caucus

On August 4, 12 of the 28 Republican state legislative incumbents facing primary challenges in Kansas lost their primaries. These results could alter the makeup of the Republican caucus in 2021.

Local media sources like The Wichita Eagle, Shawnee Mission Post, and The Kansas City Star identified nine incumbents whose primaries and defeats were representative of an intra-party divide over issues including Medicaid expansion and abortion laws.

Michael Ryan wrote in The Kansas City Star, “A good number of conservatives absolutely washed over their more moderate state legislative opponents in Tuesday’s Republican primary election.”

Jonathan Shorman wrote in The Wichita Eagle, “The influence of Kansas Republican moderates has waxed and waned. Gov. Sam Brownback [(R)] helped oust them in 2012. Voters then swept them back into office in 2016 to end his signature income tax cuts and stabilize the budget.” Shorman continued, “But with last week’s primary losses, their ranks have been depleted to levels not seen for years.”

The following six state senators lost Republican primaries this year. All were first elected in 2016, the year of the most recent state senate elections. Sens. John Skubal, Bruce Givens, Randall Hardy, and Edward Berger defeated Republican incumbents in primaries themselves that year.

• District 11: Sen. John Skubal, lost to Kellie Warren 64-36%
• District 14: Sen. Bruce Givens, lost to Michael Fagg 54-46%
• District 15: Sen. Dan Goddard, lost to Virgil Peck, Jr. 50.1-49.9%
• District 24: Sen. Randall Hardy, lost to J.R. Claeys 63-37%
• District 33: Sen. Mary Jo Taylor, lost to Alicia Straub 60-40%
• District 34: Sen. Edward Berger, lost to Mark Steffen 57.5%-42.5%

The following three state representatives lost Republican primaries this year.

• District 20: Rep. Jan Kessinger, lost to Jane Dirks 57-43%
• District 42: Rep. Jim Karleskint, lost to Lance Neelly 52-48%
• District 71: Rep. Diana Dierks, lost to Steven Howe 62-38%

Local media outlets identified the No Right to Abortion in Constitution Amendment as a noteworthy issue in the primaries featuring Sen. Skubal and Rep. Kessinger. If passed, the measure would have placed a legislatively referred constitutional amendment on the August 4 ballot stating that there is no right to abortion or public funding for abortions in the Kansas Constitution.

A two-thirds vote of all members in each chamber of the Kansas State Legislature was required to refer the amendment to the ballot for voter consideration.

In the Senate, that equaled 27 votes and, in the House, 84. Republicans held supermajorities in both the Senate and House with 29 and 84 members, respectively, equal or greater to the two-thirds threshold in both chambers.

Skubal and Kessinger voted against placing the measure on the ballot. The Senate approved the amendment. Kessinger and three other Republicans joined 39 Democrats to vote against the amendment in the House, resulting in a final vote of 80-43, four votes short of passage.

Skubal’s and Kessinger’s primary defeats and the retirement of the three House Republicans who voted against the amendment with Kessinger means none of the five Republicans who opposed the No Right to Abortion in Constitution Amendment will return for the 2021 legislative session.

Heading into the November elections, Kansas has a divided government. Republicans control both the Senate and the House. Democrats hold the governorship following Gov. Laura Kelly’s election in 2018. The next gubernatorial election will be held on November 8, 2022. In the Senate, Republicans will retain a veto-proof supermajority if they gain seats, hold their current number of seats, or lose a net of two seats. In the House, Republicans will retain a veto-proof majority if they gain or hold their current number seats. If Democrats gain a net of one seat, they will control more than one-third of the House seats.

To read more background, local commentary, and historical data on Kansas’ state legislative Republican primaries, click here.